ARTICLE IN PRESS

Atmospheric Environment 41 (2007) 1053–1063

www.elsevier.com/locate/atmosenv

Formation of fine particles enriched by V and Ni from

heavy oil combustion: Anthropogenic sources and
drop-tube furnace experiments
Ha-Na Janga, Yong-Chil Seoa, Ju-Hyung Leea, Kyu-Won Hwanga, Jong-Ik Yoob,
Chong-Hui Sokc, Seong-Heon Kima,d,
a
Department of Environmental Engineering, YIEST, Yonsei University, Republic of Korea
b
Air Pollution Prevention and Control Division, RTP, NC, US EPA
c
LG Chem/Research Park, Republic of Korea
d
Air and Waste Engineering Laboratory, Department of Environmental Engineering, #305 Back-Un Building, Yonsei University,
Heung-up, Won-Ju, Gang-Won 220-710, Republic of Korea
Received 5 May 2006; received in revised form 11 July 2006; accepted 8 September 2006

Abstract

The present study attempts to investigate the emission characteristics of fine particles with special emphasis on nickel
and vanadium metal elements emitted from the heavy oil combustion in industrial boilers and power plant, which are
typical anthropogenic sources in Korea. A series of combustion experiments were performed to investigate the emission
characteristics of particles in the size range of submicron by means of drop-tube furnace with three major domestic heavy
oils. Cascade impactors were utilized to determine the size distribution of particulates as well as to analyze the partitioning
enrichment of vanadium and nickel in various size ranges. Experimental results were compared with field data of particle
size distribution and metal partitioning at commercial utility boilers with heavy oil combustion. Such data were interpreted
by chemical equilibrium and particle growth mechanism by means of computational models. In general, fine particles were
the major portion of PM10 emitted from the heavy oil combustion, with significant fraction of ultra-fine particles. The
formation of ultra-fine particles through nucleation/condensation/coagulation from heavy oil combustion was confirmed
by field and experimental data. Vanadium and nickel were more enriched in fine particles, particularly in ultra-fine
particles. The conventional air pollution devices showed inefficient capability to remove ultra-fine particles enriched with
hazardous transition metal elements such as vanadium and nickel.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Combustion; Heavy oil; Vanadium; Nickel; Ultra-fine particles

Corresponding author. Air and Waste Engineering Labora-
tory, Department of Environmental Engineering, #305 Back-Un
Building, Yonsei University, Heung-up, Won-Ju, Gang-Won
220-710, Republic of Korea. Tel.: +82 33 760 2380;
fax: +82 33 763 5224.
E-mail address: seongheo@yonsei.ac.kr (S.-H. Kim).
1. Introduction
Airborne particulate matter has been the
primary focus of several epidemiological studies,
which reported a correlation between adverse
health effects and the particle concentration levels

1352-2310/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.atmosenv.2006.09.011
 ARTICLE IN PRESS
1054 H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063
(Dockery et al., 1993; Ozkaynak et al., 1993). US
Environmental Protection Agency (EPA) estab-
lished National Ambient Air Quality Standards
(NAAQS) for the six major criteria air pollutants,
among which particulate matters less than 10 mm
(PM10) and particulate matters less than 2.5 mm
(PM2.5) in aerodynamic diameter are associated
with adverse health effects such as asthma, respira-
tory disease, and mortality (US EPA, 1997). World
Health Organization (WHO regional office for
Europe) reported various exposure studies designed
for the evaluation of toxic effects associated with
physico-chemical properties of fine particulates
(WHO Europe, 2000). Recent epidemiological
studies have also indicated a strong correlation
between lung cancer-induced mortality and the
concentration level of fine particulates (Burnett
et al., 2001; Pope et al., 2002; Krewski et al.,
2005). Both the size and the chemical composition
of the particulates are attributed to the adverse
health effects. Adverse health effects of particulates
becomes severer as particulates are enriched by
hazardous metal elements such as vanadium and
nickel, which are usually originated from the fuel
combustion of heavy oils (Campen et al., 2001;
Kodavanti et al., 2001).
Heavy oil combustion facilities produce fine
particles in the size range of submicron, which could
be enriched by heavy metals such as vanadium, nickel
and zinc contained in liquid fuel. The very small size
of these ultra-fine particulates makes it difficult to
remove them by existing air pollution control devices
from the flue gas. Industrial boilers and electricity
generation boilers consuming heavy oil are regarded
as one of the main sources of such fine particulate
pollutants, and cause adverse health effect by emitting
particulates enriched with heavy metals into the
atmosphere. Some advanced countries have proved
hazardousness of nickel and vanadium in heavy oil
and the concentration of these elements have been
limited by emission standard. Particulates from
combustion process are formed when inorganic
compounds and metals in fuel evaporate at high
temperature and then condensate or coagulate. Linak
et al. (2000, 2003, 2004) have reported a series of
studies on the size distribution and chemical proper-
ties of fine particles formed in the combustion process
of heavy oil. According to Linak et al., the combus-
tion of heavy oils resulted in relatively pronounced
fraction of ultra-fine particles smaller than 0.1 mm, in
the size range of which transition metals such as
nickel and vanadium were mostly enriched. Anthro-
pogenic sources such as heavy oil combustion
facilities or electricity generation plants release fine
particles containing mainly vanadium and nickel, due
to the limited removal efficiency of the existing air
pollutants control devices (APCD) for these small
particles (Lighty et al., 2000). Fine particles are
thought to be originated from the metal vapor at high
combustion temperature and formed via nucleation,
condensation, coagulation processes (Linak and
Wendt, 1993). The temperature of the furnace and
the residence time were the controlling factors and the
particle size distribution depended on the interaction
between chemical reactions, nucleation, condensation,
and coagulation (Biswas and Wu, 1997).
There is a strong relationship between the size
distribution and the chemical properties of atmo-
spheric particulates mainly originated from such an
anthropogenic source. Shaheen et al. (2005) carried
out a measurement of the atmospheric concentra-
tion of 10 heavy metals (Na, K, Fe, Zn, Pb, Mn, Cr,
Co, Ni, and Cd) in the four size ranges (o2.5 mm,
2.5–10 mm, 10–100 mm, 4100 mm), and reported
that transition metals such as nickel existed mainly
in fine particle mode (o2.5 mm) and coarse mode
(2.5–10 mm) of airborne particulates. Espinosa et al.
(2001), who measured the size distribution of total
suspended particles (TSP) and heavy metals in
Spain, reported that PM10, PM2.5, PM0.61 consti-
tuted 85%, 61%, 50% of TSP, respectively. More
than 60% of Ni, V, Pb, Cd, Pb, and Cd among the
11 metals analyzed were contained in ultra-fine
particles less than 0.61 mm in aerodynamic diameter.
According to Singh et al. (2002), particulate matter
in the size range of 1–2.5 mm were constituted
mainly by organic carbons, heavy metals, nitrate,
and sulfate. And the smaller particle size was more
associated with heavy metals such as Pb, Sn, Ni, Cr,
V, 70–85% of such metals were distributed in the
submicron diameter and 40% of which in particle
size less than 0.35 mm. The strong correlation
between the size distribution of airborne particles
and heavy metal contents was attributed to the
anthropogenic source emission (Espinosa et al.,
2001; Singh et al., 2002). According to source
apportionment study by Vallius et al. (2003), heavy
oil combustion contributed 13% of PM2.5 and
vanadium, nickel, and SO2 were recognized as index
materials for heavy oil combustion. Li et al. (2004)
reported that PM2.5 in New York City was appor-
tioned to six major factors of heavy oil combustion,
automobile, suspended particulates, sea salts, and
two secondary sources as sulfate and nitrate.
 ARTICLE IN PRESS
H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063
Fuel combustion has been known to make a
significant contribution to PM2.5, particularly to
ultra-fine particles which explains most of number
concentration of atmospheric aerosols (Lighty et al.,
2000) with potential adverse health effects. There-
fore, it has become important to have a better
understanding of the emission characteristics and
the formation mechanism of ultra-fine particles.
The objective of this study was to investigate the
physico-chemical properties of mainly ultra-fine
particles emitted from heavy oil combustion with
special focus on vanadium and nickel. The size
distribution of particulate matter and the contained
hazardous metal elements were determined for the
anthropogenic sources including industrial boilers
and power plant. A laboratory combustion experi-
ment in drop-tube furnace was performed using
three major domestic heavy oils to confirm such
metal enrichment in ultra-fine particulates which
have been observed in case of industrial boilers and
power plant. Model simulation was carried out to
predict the chemical equilibrium and particle
growth by means of CEA and MAEROS codes.
The results thus obtained were compared and used
to support the experimental results from laboratory
and commercial combustion processes.
2. Facilities and experimental methods
2.1. Lab-scale drop-tube furnace
A drop-tube furnace was designed for the
laboratory experiment under the similar condition
in temperature and residence time as those of
industrial boiler and power plant. The drop-tube
furnace consists of alumina tube as heating body, in
the center of which fuel injector and air jet gun were
located. The burden capacity was up to 25 kW with
the temperature control range from 25 to 1550 1C. A
schematic diagram of the drop-tube furnace utilized
in the present study is depicted in Fig. 1. Fuel oil
was injected by a cartridge pump (Model No
7521–50, Col-Parmer Inc.) with the rate of
3–5 g h 1. The combustion experiment was per-
formed at 1400 1C followed by two cooling steps
through water circulation and air dilution.
2.2. Facilities tested for the anthropogenic sources
The facilities, tested for the anthropogenic source
combustion in this study, were two boilers (Oil_ind
1, Oil_ind 2) and 1 power plant (Oil_pwr). Two
1055
boilers utilized bunker-C oil (S content: 40.3%)
with steam generation capacity of 4.5 and 10
ton h 1, respectively. The tested power plant utilized
bunker-C oil (0.3% of sulfur content) with steam
generation capacity of 1135 ton h 1 (350 MW in
electricity). Sampling probe was located at the stack
point after the cyclone, the air pollution control
device of industrial boilers. Another sampling probe
was located before the electrostatic precipitator
(ESP) in addition to the stack point after ESP in
case of power plant. Detailed information of
facilities is listed in Table 1.
2.3. Sampling and analysis
A cascade impactor (Anderson Instrument Co.
Ltd.) was utilized for the sampling and the
determination of size fractionized mass concentra-
tion of particulate matter emitted from anthropo-
genic oil combustion. Iso-kinetic coefficient was
maintained in the range of 95–110%. Table 2 shows
the ultimate compositions of three domestic heavy
oils tested in the drop-tube furnace. The tested three
types of bunker-C oils were characterized by
relatively high fraction of vanadium and nickel as
well as high sulfur content.
Micro-Orifice Uniform Deposit Impactor
(MOUDI, MSP Co. Ltd., Minneapolis, MN) was
utilized for the sampling and the determination of
size segregated mass concentration of particulates
emitted from the laboratory combustion experi-
ment. MOUDI impactor consists of ten stages with
the 50% collection efficiency cut-points in the range
of 0.05–18 mm under the operating flow rate of
30 L min 1. For both field and lab scale tests,
sampling process was followed to EPA method
201A—Determination of PM10 emission and size
distribution (US EPA Method, 1997) and was
repeated three times. Metal elements in particulates
were analyzed by ICP/MS (Varian Co. Ltd., Ultra
mass 700) after pretreatment of samples as
documented in EPA Method 3050B (US EPA
Method, 1986).
3. Result and discussion
3.1. Emission characteristics of PM10 emitted from
the laboratory drop-tube furnace
Particle size distribution (PSD) of PM10 emitted
from laboratory drop-tube furnace is shown in
Fig. 2. The highest mass fraction was observed
 ARTICLE IN PRESS
1056 H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063

Oil Inlet

Water Water
A B inlet outlet

Pump

B-Coil
Water bath

C
D

Main controller
A : Oil inlet probe
B : Thermo couple
C : Alumina Furnace Tube
D : Heating area
E : Sampling probe air
F : Pressure gauge E
G : Pump
H : Water controller

H
Water Water
G
MOUDI (Filter) inlet outlet
Air Air

Sample port

Fig. 1. Schematic diagram of lab-scale drop-tube furnace.

around 0.1 mm within the ultra-fine particle size
range. More than 90% of PM10 existed in the size
range less than 3.1 mm (PM2.5), 80% of which
existed within PM0.5. Especially, the averaged
mass fraction of PM0.1 consisted of more than
50% of PM10, 60% of PM2.5, and 67% of PM0.5,
respectively. The above results indicated that
particulate matter emitted from the heavy oil
combustion mainly consisted of fine particulates
less than 2.5 mm in aerodynamic diameter, with
dominating mass fraction in the size range of ultra-
fine particles.
 ARTICLE IN PRESS
H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063 1057

Table 1
Identification and capacity for tested boilers

ID Capacity (steam generation) Fuel Boiler type APCDa

Oil_ind 1 4.5 ton h 1 B/B oil Fire tube Cyclone

Oil_ind 2 10 ton h 1 B/C oil Fire tube Cyclone
1
Oil_pwr 1135 ton h (350 MW) B/C oil Electric generation ESP
a
APCD denotes Air Pollutants Control Device.

Table 2
Ultimate analysis of heavy oil used in lab-scale oil combustion

Oil-1 Oil-2 Oil-3

Water (%) 0.03 0.01 0.05

Carbon (%) 83.15 82.72 85.63
Hydrogen (%) 9.79 10.52 11.09
Nitrogen (%) 0.76 1.18 0.81
Sulfur (%) 1.65 3.02 1.23
Ash (%) 5.67 3.92 5.78
Oxygen (%) 0.96 0.91 0.86
Vanadium (mg g 1) 38.70 58.80 40.87
Nickel (mg g 1) 18.50 22.40 17.57
Iron (mg g 1) — — —
Zinc (mg g 1) 56.80 — —

Fig. 2. Particle size distribution of PM10 emitted from lab-scale

3.2. Emission characteristics of PM10 emitted from combustion.
oil combustion facilities

Particle size distributions (PSD) of PM10 emitted

from the heavy oil combustion facilities are shown
in Fig. 3. On average, 23% of PM10 mass
concentration at the stack point was contained
within ultra-fine particle size range (o0.1 mm), and
80% within 2.5 mm. As shown in Fig. 3, bi-modal
size distribution was dominant except for Oil_pwr
stack. The formation of fine mode around 1 mm was
explained by either incomplete burnout of oil or
residuals of inherent cenospheres contained in oil
(Miller et al., 1998; Linak et al., 2000). Single mode
in the size range of ultra-fine particles observed in
case of Oil_pwr stack can be interpreted to come
from vaporization and following condensation
processes of trace metals, while particles larger than
Fig. 3. Particle size distribution of PM10 emitted from the oil
0.3 mm are efficiently removed by Electrostatic combustion facilities.
Precipitator (ESP). The average mass concentration
of PM10 was decreased from 6.4 mg m 3 before the
ESP to 0.4 mg m 3 after the ESP. coagulation processes for the residence time be-
The submicron mode (o1 mm) observed before tween the boiler and the stack. As confirmed in
APCD in power plant is related to further aging laboratory experiment by means of drop-tube
process of ultra-fine particles formed from metal furnace combustion, ultra-fine mode within 0.1 mm
vapors. Metal vapor can undergo condensational is formed via vaporization/condensation processes
nucleation and then grow via condensation and of trace metals such as V and Ni, and then grow to

1058
fine mode up to 1 mm by coagulation process. In
order to validate the assumption that the ultra-fine
particle are formed by nucleation, condensation,
coagulation processes from the evaporated metal
vapors, laboratory experiments using drop-tube
furnace were performed and the results were
compared with the computational calculation by
MAEROS code as described in following section.
3.3. Model prediction for particle growth
In order to illustrate particle growth by nuclea-
tion, condensation, and coagulation, MAEROS was
used to simulate particle growth (Gelbard and
Seinfeld, 1980). Initial concentration was given as
14 mg m 3, the measured value in drop-tube furnace
according to EPA method 5. The result from lab-
scale combustion of Oil-1 was taken as input
data used in model prediction. Primary particle
diameter of 2 nm was allocated to the Section 2
(0.0021–0.0046 mm) in the simulation process for
nucleation. Pressure was maintained to 1.01 
105 Pa, while temperature was ramped in 0.1 s from
the tube temperature of 1650 K down to 373 K
which is the sampling probe temperature adjusted
by EPA method 5. The initial number concentration
was 1.1  1017 m 3. Fig. 4 shows the predicted
particle size distribution after 0.1, 0.5, 1, 2, 5, and
10 s of residence time. As time increased from 0.1 to
10 s, number concentration decreased to 8.3  1016,
4.9  1015, 9.2  1014, 1.9  1014, 2.5  1014, and
6.0  1012 m 3, respectively. As shown in Fig. 4,
the accumulation mode around 0.05 mm appeared
20
Mass Concentration (mg/m3)
15
10
0
0.001
ARTICLE IN PRESS
H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063
after 5 s of residence time, which was consistent with
the experimental result of Oil-1 combustion. Fig. 4
also showed that the accumulation mode started to
diminish distinctly after 10 s of residence time
indicating the difficulty for particles to grow larger
than 1 mm only by the coagulation mechanism. The
above model calculation implied that the accumula-
tion mode around 0.3 mm observed before APCD in
Oil_pwr was the result of further growth of ultra-
fine particles after longer residence time.
3.4. Evaporation prediction by chemical equilibrium
The chemical equilibrium analysis (CEA) code
(McBride et al., 1993) was used to predict the
element speciation and, more importantly, the
fractions of vapor-phase species as a function of
temperature for principle metals. References for the
thermo-chemical properties and species data sets
used are found in Linak et al. (2003). The predicted
compound speciation of nickel and vanadium at
various temperature ranges are shown in Fig. 5.
Data from the US EPA study (Linak et al., 2004)
were depicted also for the comparison. Both
vanadium and nickel exist in the vapor phase at
the temperature range higher than 1400 1C. As
shown in Fig. 5, the combustion temperature of
1600 K inside drop-tube furnace was high enough to
vaporize V and Ni. V and Ni are predicted to form
sulfates at lower temperature (400–900 K) and
oxides at higher temperature (higher than 1000 K).
Sulfate formation affects the solubility and perhaps
0.1sec
0.5sec
1sec
2sec
5sec
10sec
0.01 0.1 1 10
Aerodynamic Diameter (µm)
Fig. 4. Model prediction by MAEROS.
 ARTICLE IN PRESS
H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063 1059

1.0 1.0
Fraction of Each Compound

Fraction of Each Compound

0.8 0.8
*Ni *Ni
0.6 NiO 0.6 NiO
NiO2H2 NiO2H2
0.4 NiSO4(S) 0.4 NiSO4(S)
NiO(3) NiO(3)
0.2 NiS 0.2 NiS

0.0 0.0
400 800 1200 1600 2000 2400 400 800 1200 1600 2000 2400
(a) Temperature (K) (b) Temperature (K)

1.0 1.0
Fraction of Each Compound

Fraction of Each Compound

0.8 0.8
*V
*Ni
0.6 0.6 *VO
NiO
VO2
NiO2H2
0.4 NiSO4(S)
0.4 V2O4(I)
V2O5(L)
NiO(3)
0.2 0.2 VOSO4(S)
NiS

0.0 0.0
400 800 1200 1600 2000 2400 400 800 1200 1600 2000 2400
(c) Temperature (K) (d) Temperature (K)

1.0 1.0
Fraction of Each Compound

Fraction of Each Compound

0.8 *V 0.8 *V
*VO *VO
0.6 VO2
0.6
VO2
V2O4(I) V2O4(I)
0.4 0.4
V2O5(L) V2O5(L)
VOSO4(S) VOSO4(S)
0.2 0.2

0.0 0.0
400 800 1200 1600 2000 2400 400 800 1200 1600 2000 2400
(e) Temperature (K) (f) Temperature (K)
1.0 1.0
Fraction of Each Compound

Fraction of Each Compound

0.8 0.8
Ni
NiO VO2
0.6 Ni(OH)2 0.6 V2O4(II)
NiSO4(s) V2O5(s)
0.4 NiFe2O4(2) 0.4 V2O5(L)
NiO(3) Condensed
Condensed phase
0.2 phase 0.2

0.0 0.0
400 800 1200 1600 2000 2400 400 800 1200 1600 2000 2400
(g) Temperature (K) (h) Temperature (K)

Fig. 5. Speciation of vanadium and nickel compounds from Oil-1, Oil-2 and Oil-3 depending on temperature variation. (a) Ni from Oil-1,
(b) Ni from Oil-2, (c) Ni from Oil-3, (d) V from Oil-1, (e) V from Oil-2, (f) V from Oil-3, (g) Ni from EPA, (h) V from EPA.
 ARTICLE IN PRESS
1060 H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063
the bioavailability and toxicity associated with V,
Ni (Linak et al., 2000).
3.5. Metal element analysis
Fig. 6 shows the size segregated mass fraction of
nickel in particulates emitted from the anthropo-
genic source. The averaged mass fraction of nickel
in PM0.5 was considerably high around 63% as
shown in Fig. 6. The mass fraction of nickel in
PM0.5 (90%) was much higher at the stack point of
Oil_pwr than the value (50%) measured before the
APCD. On the contrary, the mass fraction of nickel
in coarse particles (2.5–10 mm) did not show a
significant change between sampling points before
APCD and at the stack point. This result of higher
enrichment of nickel in fine mode indicated that fine
particles were grown as a result of metal evapora-
tion followed by nucleation, condensation, and
coagulation. A similar enrichment pattern of metal
elements was observed in the laboratory combus-
tion experiment. The mass fraction distribution for
both vanadium and nickel showed the peak value in
ultra-fine particle size range as depicted in Figs. 7
and 8. The mass fractions of nickel and vanadium in
fine particles were 66% and 74%, respectively. And
the ratio of nickel content in PM0.1 to that in PM2.5
was 48% and the ratio of vanadium content in
PM0.1 to that in PM2.5 was 53%. It could clearly
confirm that metals were vaporized at high combus-
tion temperatures as predicted by chemical equili-
brium code, then formed ultra-fine particulates by a
100
80
Mass Fraction (%)
60
40
20
0
< 0.5
Aerodynamic Diameter (µm)
Fig. 6. Size segregated mass fraction of nickel in PM10 from anthropogenic oil combustion sources.
suggested particle growth model up to the size under
1 mm.
The enrichment portion of metals and particle
size distribution were compared between lab-scale
furnace and industrial oil boilers as summarized in
Table 3. Both data from lab-scale furnace and real
industrial boilers for oil combustion showed similar
pattern with high enrichment of metals in ultra-fine
particulates. The portion of metal enriched in fines
from field experiment was less than that from
laboratory experiment.
4. Conclusions
The emission characteristics of fine particles was
investigated with special emphasis on nickel and
vanadium metal elements emitted from the heavy oil
combustion in industrial boilers and power plant,
which are typical anthropogenic sources in Korea.
PM10 emitted from the facilities mainly consisted of
fine particles (80%), which was dominated by ultra-
fine particles (23%), on average. Particularly in
power plant, the mass fraction of ultra-fine particles
(PM0.1/PM10) was greatly enhanced during the
transfer from 1% measured before APCD to 40%
at the stack point. The averaged mass fraction of
nickel in PM0.5 also showed considerably high value
of 63%. The ratio of nickel fraction in ultra-fine
particles to that in PM10 was increased almost twice
during transfer from before APCD to the stack of
the oil power plant. The implication of the field
results is that conventional control device might be
Oil_Ind1 stack
Oil_Ind2 stack
Oil-Pwr(before APCD)
Oil_Pwr stack
Average(stack)
2.5-0.5 10-2.5
 ARTICLE IN PRESS
H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063 1061

60
Oil 1
Oil 2
Oil 3
50
Avg.

40
Mass Fraction (%)

30

20

10

0
<0.1 0.1-0.18 0.18-0.31 0.31-0.51 0.51-1.0 1.0-1.8 1.8-3.1 3.1-6.2 6.2-10
Aerodynamic Diameter (µm)

Fig. 7. Size segregated mass fraction of nickel in PM10 from lab-scale oil combustion.

70
Oil 1
Oil 2
60 Oil 3
Avg.

50
Mass Fraction (%)

40

30

20

10

0
<0.1 0.1-0.18 0.18-0.31 0.31-0.51 0.51-1.0 1.0-1.8 1.8-3.1 3.1-6.2 6.2-10
Aerodynamic Diameter (µm)

Fig. 8. Size segregated mass fraction of vanadium in PM10 from lab-scale oil combustion.

ineffective for the removal of ultra-fine particles
being grown during the transfer by nucleation,
condensation, and coagulation of metal vapor. The
above results were verified by laboratory combustion
experiments in drop-tube furnace and model simula-
tion using MAEROS code. The primary nano-size
particles were predicted by MAEROS computation
to grow in 2 s to ultra-fine particle size of 0.1 mm,
which was in a good agreement with experimental
results from lab-scale drop-tube furnace. Ultra-fine
particles consisted more than 50% of PM10 from the
lab-scale oil combustion. Furthermore, the enrich-
ment of nickel and vanadium elements was more
pronounced in the smaller particle size, with the peak
value at 0.1 mm. It was difficult for the conventional
air pollution devices to remove ultra-fine particles
enriched with hazardous transition metal elements
such as vanadium and nickel, which would raise a
request for new establishment or the revise of
emission regulatory standards.
 ARTICLE IN PRESS
1062 H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063

Table 3 Kodavanti, U.P., Schladweiler, M.C.J., Richards, J.R., Coata,

Comparison of PSD in PM10 and mass fraction of metals in D.L., 2001. Acute lung injury from intratracheal exposure to
different size ranges between lab-scale experiment and industrial fugitive residual oil fly ash and its constituent metals in
boilers normo- and spontaneously hypertensive rats. Inhalation
Toxicology 13 (1), 37–54.
Lab-scale data Field data Krewski, D., Burnett, R., Jerrett, M., Pope, C.A., Rainham, D.,
Calle, E., Thurston, G., Thun, M., 2005. Mortality and long-
PM10 PSD (%) term exposure to ambient air pollution: ongoing analyses
o2.5 mm 90.49 79.53 based on the American cancer society cohort. Journal
o0.5 mm 81.60 54.10 of Toxicology and Environmental Health, Part A 68,
o0.1 mm 54.35 22.70 1093–1109.
Nickel MF (%) Li, Z., Hopke, P.K., Husain, L., Qureshi, S., Dutkiewicz, V.A.,
o2.5 mm 90.04 85.24 Schwab, J.J., Drewnick, F., Demerjian, K.L., 2004. Sources
o0.5 mm 66.35 63.20 of fine particle composition in New York city. Atmospheric
o0.1 mm 43.21 41.16a Environment 38, 6521–6529.
Lighty, J.S., Vernath, J.M., Sarofim, A.F., 2000. Combustion
Vanadium MF (%) aerosols : factors governing their size and compostion and
o2.5 mm 91.25 — implications to human health. Journal of the Air & Waste
o0.5 mm 74.05 — Management Association 50, 1565–1618.
o0.1 mm 48.33 — Linak, W.P., Wendt, J.O.L., 1993. Toxic metal emissions from
a
incineration: Mechanisms and control. Progress in Energy
The prediction value of Ni mass fraction in PM0.1 by the Combustion Science 19, 145–185.
combustion of drop-tube furnace (%). Linak, W.P., Miller, C.A., Wendt, J.O.L., 2000. Fine particle
emissions from residual fuel oil combustion: characterization
and mechanisms of formation. Symposium (International) on
Acknowledgements Combustion 28 (2), 2651–2658.
Linak, W.P., Miller, C.A., Santoianni, D.A., King, C.J.,
Shinagawa, T., Wendt, J.O.L., Yoo, J.I., Seo, Y.C., 2003.
This work was supported by R&D fund from Formation of fine particles from residual oil combustion:
Korea Energy Management Corporation and Kor- reducing nuclei through the addition of inorganic sorbent.
ea Institute of Environmental Science and Technol- Korean Journal of Chemical Engineering 20, 664–669.
ogy. Model calculations and analysis were assisted Linak, W.P., Miller, C.A., Wood, J.P., Shinagawa, T., Yoo, J.I.,
by W.P. Linak in US EPA. Santoianni, D.A., King, C.J., Seo, Y.C., 2004. High
temperature interactions between residual residual oil ash
and dispersed kaolinite powders. Aerosol Science and
Technology 38 (9), 900–913.
References McBride, B.J., Gordon, S., Reno, M.A., 1993. Coefficients for
calculating thermodynamic and transport properties of
Biswas, P., Wu, Y.C., 1997. Control of toxic metal emission from individual species. NASA, Technical Memorandum 4513.
combustors using sorbents. Journal of the Air & Waste Miller, C.A., Linak, W.P., King, C., Wendt, J.O.L., 1998. Fine
Management Association 48, 113–127. particle emission from heavy fuel oil combustion in a firetube
Burnett, R.T., Ma, R., Jerrett, M., Goldberg, M.S., Cakmak, S., package boiler. Combustion Science and Technology 134,
Pope III, C.A., Krewski, D., 2001. The spatial association 477–502.
between community air pollution and mortality: a new Ozkaynak, H., Xue, J., Severance, P., Bunnet, R., Raizeene, M.,
method of analyzing correlated geographic cohort data. 1993. Association between daily mortality, ozone, and
Environmental Health Perspectives 109 (3), 375–380. particulate air pollution in Toronto, Canada. Presented at
Campen, M.J., Nolan, J.P., Schladweiler, M.C.J., Kodavanti, the Colloquium on Particulate Air Pollution and Human
U.P., Evansky, P.A., Costa, D.L., Watkinson, W.P., 2001. Mortality and Morbidity, Irvine, CA, 24–25 January.
Cardiovascular and thermoregulatory effects of inhaled PM Pope III, C.A., Burnett, R.T., Thun, M.J., Calle, E.E., Krewski,
associated transition metals: a potential interaction between D., Ito, K., Thurston, G.D., 2002. Journal of the American
nickel and vanadium sulfate. Toxicological Sciences 64 (2), Medical Association 287 (9), 1132–1141.
243–252. Shaheen, N., Shah, M.H., Jaffar, M., 2005. A study of airborne
Dockery, D.W., Pope, A.C., Xu, X., Spengler, J.D., Ware, J.H., selected metals and particle size distribution in relation to
Fay, M.E., Ferris, B.G., Speizer, F.E., 1993. An association climatic variables and their source identification. Water, Air,
between air pollution and mortality in size US cities. & Soil Pollution 164, 275–294.
New England Journal of Medicine 329, 1753–1759. Singh, M., Jaques, P.A., Sioutas, C., 2002. Size distribution and
Espinosa, A.J.F., Rodriguez, M.T., Rosa, F.J., Sanchez, J.C.J., diurnal characteristics of particle-bound metals in source and
2001. Size distribution of metals in urban aerosols in Seville receptor site of the Los Angeles Basin. Atmospheric
(Spain). Atmospheric Environment 35, 2595–2601. Environment 36, 1675–1689.
Gelbard, F., Seinfeld, J.H., 1980. Simulation of multicomponent US Environmental Protection Agency, 1997. Technology Trans-
aerosol dynamics. Journal of Colloid and Interface Science fer Network OAR Policy and Guidance. /http://www.epa.
78, 485–501. gov/ttn/oarpg/naaqsfin/pmhealth.htmlS
 ARTICLE IN PRESS
H.-N. Jang et al. / Atmospheric Environment 41 (2007) 1053–1063 1063

US Environmental Protection Agency, 1997. CFR Promulgated
Test Methods. /http://www.epa.gov/ttn/emc/promgate.htmlS
US Environmental Protection Agency, 1986. Test Methods SW-
846. /http://www.epa.gov/epaoswer/hazwaste/test/sw846.htmS
Vallius, M., Lanki, T., Tiittanen, P., Koistinen, K., Ruuskanen,
J., Pekkanen, J., 2003. Source apportionment of urban
ambient PM2.5 in two successive measurement campaigns
in Helsinki, Finland. Atmospheric Environment 37,
615–623.
WHO Europe, 2000. Air quality guidelines second edition
Regional Office for Europe. /http://www.euro.who.int/air/
activitiesS